JPH0452981B2 - - Google Patents

Description

[Detailed Description of the Invention] [Summary] When starting a job on the local system from the host system, instead of starting the job that was started first by the host system and then starting the next job. A completion code indicating the status of the job is stored separately, and the job is activated continuously for the local system to improve control efficiency.

[Industrial application field]

The present invention relates to a method for controlling jobs between systems in a data processing system consisting of a host system and a local system, and more particularly to a control method for continuously starting jobs in a local system from a host system.

[Conventional technology]

There are two methods for starting jobs on the local system from the host system: (a) push-off type and (b) wait-for-completion type.

In the push-out type (a), the host system requests the local system to start a job, and then the host system has no involvement in the local system's jobs.

(b) Wait-for-completion type is one in which the host system starts a job on the local system, waits for the job on the local system to finish, and performs the next process after receiving the job completion report.

Conventionally, when starting a job on the local system from the host system in a wait-for-completion type, the series of operations was completed when the processing program that issued the request to start the host system received a report on the completion of the job on the local system. .

FIG. 3 shows the processing procedure of conventional job control using the wait-for-completion type for both the host system and the local system. A brief explanation follows.

A processing program is started on the host system.

A local system job activation request occurs in the processing program.

The job launches successfully on the local system.

The job on the host system waits for completion (WAIT).

Perform processing on the local system.

The local processing job ends.

Completion is reported from the local side to the host system.

Receive the completion report in the asynchronous module and release the wait (POST).

The host processing program receives the completion code from the local job and processes it.

The processing program ends.

[Problem that the invention seeks to solve]

In conventional systems, when a normal job in the host system issues a job startup request to the local system, the operator must press the interrupt key to perform other processing using the time while waiting for the completion report. It is often done to take advantage of the wait time to execute an interrupt job. Note that, in general, the priority of interrupt jobs is set higher than that of normal jobs.

Now, if this interrupt job starts a job again on the local system, even if the local job that the normal job previously requested has finished, the sequence of local job starts by the normal job on the host system will continue. If the operation is not yet completed, the host system OS
had to suppress local job activation requests for interrupt jobs.

FIG. 4 shows a processing procedure when an interrupt job is executed during the waiting time after a local job is activated by a normal job in such a host system. A brief explanation follows.

Start a normal job on the host system.

The local job startup command is executed.

Confirm that the local job ECB 201 is free.

The state of the local job ECB 201 is set to the waiting state.

Sends a local job startup request to the local system.

A local job launch occurs on the local system.

Normal jobs on the host system go into WAIT state.

The local system performs local job processing.

An interrupt is made on the host system using the interrupt key, causing an interruption.

An interrupt job is started.

The local job completes, and the local system reports completion and a completion code to the host system.

The asynchronous processing routine of the host system that receives the completion of the local job sets the local job ECB 201 in the waiting state to the completed state and sets the completion code to 202.

A local job start command is executed in an interrupt job.

It is known from the local job ECB 201 that the normal job is using the local job ECB.

Outputs an error message when local job startup fails.

The operator sees the message and cannot start the local job, so he cancels the interrupt job.

Since the interrupt job has ended, the normal job is resumed.

Normal job is by ECB for local job.
Know when a local job is completed.

The normal job copies the completion code to the normal job's completion code storage area 202, and makes the local job ECB free.

A normal job ends with the value of the normal job completion code storage area.

In this way, when an interrupt job in the host system requests a local job, it may be necessary to cancel the interrupt job, have the job receive a completion report, and then re-execute it, which can cause problems in processing efficiency. It was hot.

In other words, when a job on the host system successively starts jobs on the local system, for example, while host job A is starting local job a, if host job B, which has a higher execution priority on the host than host job A, starts local job b, then Even if local job a has finished and host job A has received the notification, host job A has been interrupted by host job B and is therefore unable to recognize the completion of local job a. Therefore, there was a problem in that host job B could not start the next local job b assuming that host job A's local job startup had not been completed.

[Means for solving problems]

In the conventional method, the workstation has only one event waiting control block (ECB) for receiving completion reports, so the OS receives the completion report from the host system processing program that has issued a startup request to the local system. Until you receive it,
I had to control it so that it would not accept requests to start the next local job.

Therefore, in the present invention, event wait control blocks (ECBs) are provided for each job execution priority level, such as for normal jobs and interrupt jobs, and when a job completion notification is received from the local system,
Performs a wait release (POST) to the event wait control block (ECB) of each request source,
The problem is solved by storing the completion code in the area following the ECB.

FIG. 1 is a block diagram illustrating the principles of the invention in an exemplary manner.

In the figure, 10 is the host system, 11 is the local system, 101 is the ECB for normal jobs,
103 is ECB for interrupt job, 105 is ECB for operator command, 102, 104, 10
6 represents a completion code, and 107 represents a completion report storage processing unit.

In the illustrated example, a normal job, an interrupt job,
In the order of operator commands, the priority level for job execution increases as the job execution progresses.

ECB101,1 of each character event wait control block
03 and 105 have a size of 4 bytes, for example, and are composed of 1 WAIT bit, 1 completion bit, and the remaining completion code. The completion code indicates information about the completion event.

The completion report storage processing unit 107 is an asynchronous module, and when it detects a local job completion report from the local system 11, it cancels the waiting state display of the corresponding ECB and stores a completion code.

[Effect]

In the configuration shown in FIG. 1, when a local job is started in a job that is being executed by the host system 10, a wait state indication is set in the ECB of the job that is being executed (for example, WAIT bit =
“1”, completion bit = “0”). If there is a local job completion report, the completion report storage processing unit 10
7, the wait state display of the ECB is canceled (WAIT bit = "0", completion bit = "1"), and a predetermined completion code is stored.

If a job being executed on the host system 10 starts a local job and is in the waiting state, if that job is interrupted and a higher-level job or operator command is started, a request to start the local job will occur again during the execution. If so, the next local job is started without confirming the completion of the local job by the preceding job (lower job) (of course, in this case, the local system must be in a startable state).

In other words, when a host job that is starting a local job on the host system is interrupted by a host job with a higher priority, even if a termination notification has been received from the local job that started earlier,
If the series of operations related to starting the local job has not yet been completed, the ECB and completion code are managed and processing is performed after the interrupt is recovered.
Even if multiple interrupts are performed, each job
ECBs and completion codes are managed similarly. Whether or not the local job can be started by the interrupt source host job can be easily determined by referring to the ECB and completion code managed by the interrupt destination host job.

In this way, it is possible to sequentially start local jobs using a plurality of jobs having different job execution priority levels without checking for completion of preceding local jobs.

〔Example〕

FIGS. 2A and 2B show a processing procedure of an embodiment based on the configuration of the present invention shown in FIG. 1.

In the example shown in FIG. 2, a local job start request 1 to the local system 11 occurs while the host system 10 is executing a normal job, and during the WAIT time, the host system 10 executes the interrupt job using the interrupt key. This is for the case where a local job activation request 2 to the local system occurs again, and corresponds to the conventional example shown in FIG. The following will explain steps 27 to 27 shown in FIGS. 2A and 2B.

A normal job is started on the host system 10.

The local job startup command is executed.

ECB101 for normal jobs and ECB for interrupt jobs indicate that the local job is not waiting for completion.
103. Confirm by ECB 105 for operator command.

The state of the ECB 101 for normal job is set to the waiting state.

A local job activation request is sent to the local system 11.

A local job is started in the local system 11.

Normal job on host system 10 is WAIT
state.

The local system 11 executes local job processing.

An interrupt is made in the host system 10 using an interrupt key, resulting in an interruption.

An interrupt job is started.

The local job is completed, and the local system 11 reports the completion and the completion code to the host system 10.

The asynchronous processing routine of the host system 10 that receives the completion of the local job changes the ECB waiting for the completion of the local job to ECB 101 for normal jobs and ECB 101 for interrupt jobs.
ECB103, ECB10 for operator commands
Search from 5.

ECB1 for normal job in waiting state
Set 01 to completed state and set completion code to 102
Set to .

A local job start command is executed in an interrupt job.

ECB101 for normal jobs and ECB for interrupt jobs indicate that the local job is not waiting for completion.
103, the operator command ECB 105 is checked and the state of the interrupt job ECB 103 is set to the waiting state.

A local job activation request is sent to the local system 11.

A local job is started in the local system 11.

The host system 10 interrupt job is
Enters WAIT state.

The local system 11 executes local job processing.

The local job is completed, and the local system 11 reports the completion and the completion code to the host system 10.

21 The asynchronous processing routine of the host system 10 that receives the completion of the local job changes the ECB waiting for the completion of the local job to ECB 101 for normal jobs and ECB 101 for interrupt jobs.
ECB103, ECB10 for operator commands
Search from 5.

22 for pending interrupt jobs
Set the ECB 103 to the completed state and set the completion code to 104.

23 The host system 10 restarts the process, copies the local job completion code 104 to the interrupt job completion code storage area, sets the interrupt job ECB 103 to an empty state, and ends the process based on the local job activation request 2.

24 The interrupt job ends with the value of the interrupt job completion code storage area.

25 The host system 10 resumes the normal job upon completion of the interrupt job.

26 The completion of the local job is known based on the completion status of ECB101 for normal jobs, and the completion code 1 is
02 is copied to the normal job completion code storage area, the normal job ECB 101 is set to an empty state, and the processing based on the local job activation request 1 is ended.

27 A normal job ends with the value of the normal job completion code storage area.

As described above, according to the present invention, generally in the host system 10, when any job being executed is interrupted by a job with a higher priority, that is, when a job is interrupted by an interrupt key for a normal job or by an operator command. , and when a job is executed by an operator command for an interrupt job, when both of those sequential jobs execute a local job startup, the subsequent local job startup does not require any completion confirmation for the preceding local job startup. .

〔Effect of the invention〕

In the present invention, when jobs are successively started from the host system to the local system, there is no need to confirm the completion of the preceding local job, so if the completion confirmation cannot be obtained as in the conventional case, the local job start request can be canceled and re-requested. This eliminates the need for this process, and processing efficiency is significantly improved.

[Brief explanation of drawings]

FIG. 1 is a configuration diagram showing the principle of the present invention, FIG. 2 is an explanatory diagram showing the processing procedure of one embodiment of the present invention, and FIG.
FIG. 4 is an explanatory diagram showing the processing procedure of conventional completion wait type job control, and FIG. 4 is an explanatory diagram showing the processing procedure when executing a conventional interrupt job. In Figure 1, 10... host system, 11...
Local system, 101...For normal jobs
ECB, 102... Completion code, 103... ECB for interrupt job, 104... Completion code, 1
05... ECB for operator command, 106...
...Completion code, 107...Completion report storage processing section.

Claims (1)

[Claims] 1. Host system 10 and local system 11
In a data processing system configured with the following, completion report storage means 101, 103, and 105 for each job execution level are provided in the host system 10, and when a job in the local system 11 is started, notifications are subsequently received from the local system 11. The completion report is stored in the completion report storage means 101, 10.
3,105, and after a job at a certain run level on host system 10 starts a job on local system 11,
When a job with a higher execution level than the job that started the host system 10 needs to successively start jobs in the local system, a completion code indicating the status of the job in the local system that started first is stored separately. 1. A local job startup control method characterized by starting a job in the next local system.